Achieving Ultimate End User Satisfaction

By
Chris Spacko
Contents

Introduction...................................................................................................................................3
Traditional Management Solutions...............................................................................................3
The End User ................................................................................................................................4
Detailed Component Metrics ........................................................................................................4
The Business Transactions............................................................................................................4
Service Level Tracking..................................................................................................................5
Application Performance Management ........................................................................................7
About the Author ...........................................................................................................................7
About Quest Software ...................................................................................................................7

2
Achieving Ultimate End User Satisfaction
By Chris Spacko

Introduction
In today’s high paced competitive world, companies rely on technology to increase
productivity, improve efficiency, streamline costs, and create a competitive advantage
for their business. Critical business applications supply valuable information regarding
customers, products, and other information that is vital to the success of the
organization. With this perspective, the “customer” has become not only the internal
end-users that must be satisfied with enterprise applications, but also the external
business partners and customers using supply chain applications and Web-enabled
storefronts.

In order to provide a satisfactory user experience and optimum level of service to both
these internal and external customers, complex IT infrastructures are required to
support these critical applications behind the scenes. Yet, as complex as the
infrastructure that supports the applications, in the eyes of the customer that complexity
is irrelevant. The user of the application is only concerned with the time it takes to
perform the task they are trying to accomplish, whether it’s placing an online order,
entering payroll, looking up customer information or various other tasks. And in the
end, poor application performance or, worse yet, downtime can be devastating for an
organization causing productivity to fall, an increase in expenses, and a loss in revenue.
This consequentially is the exact opposite benefit of what enterprise applications are
designed to provide. As such, it is imperative that companies proactively measure their
most critical applications to detect and resolve application problems before their
business is adversely affected.

Traditional Management Solutions

Traditional management solutions were designed with the intent to provide reactive
alerting to system faults or outages. The basic focus of these tools is to provide the IT
organization with availability coverage of the enterprise resources and infrastructure
through the use of a centralized console. Most organizations created a NOC (Network
Operation Center) or a central location of management for such tools. These
management solutions also focused heavily on scalability and being able to provide
fault tolerance alerting for an entire enterprise. This has become the issue of breadth
versus depth of information. For organizations who have implanted such a solution, this
may provide value for the enterprise, but often falls short in its ability to implement a
more proactive approach to application performance management (APM), specifically
the ability to determine the cause of performance problems before they affect your
users.

3
The End User
Using a traditional monitoring solution will leave an organization void of
understanding what is happening with their end users (customers to service), often
times leaving the telephone, e-mail or other standard method to find out when end users
are having performance problems. A complete end-to-end (E2E) performance
monitoring solution should not only provide notifications when systems fail but also
proactively monitor and alert on application trends, end user experience (application
response time), and provide detailed application performance metrics. This should be
provided by a robust solution that a manager, application owner, business analyst, or
other administrator could use to quickly determine the cause of application
performance problems and resolve the issues. Better yet, through proactive
performance monitoring, trending and alerting, an administrator may be able to head
off possible application problems before the end user experiences any degradation in
application response time.

Detailed Component Metrics

Beyond the end user, an organization must track system and application performance
metrics. An end-to-end performance monitoring solution provides two levels of
functionality: end user response time and application performance metrics.

The metrics collected should go beyond the traditional system level availability state
events. To analyze the performance of an application you need to look at the
“application stack”, which is comprised of all the components that make up the
application:

End User > Web Server > Application Server > Database

Of course this is a drastically oversimplified version of the application stack. Many

other components can be used but this illustrates a core infrastructure for an
application. Being able to see performance across the application stack is imperative to
providing the end user with the best experience possible.

The Business Transactions

End user response time measurement is used to determine how the end users are experiencing
your application. In order to gain visibility into the performance and availability of key
business processes and applications, you must determine what are the critical business
transactions you want to simulate in order to monitor the end-user experience. Determining
how an application is being utilized in your production environment, what types of business
transactions are being performed and with what frequency, can be done by interacting directly
with the end users, working with business analysts and reviewing Web logs. Once the most
utilized business transactions (repetitive) and the longest and/or problematic transactions are
identified, an administrator may be able to head off possible application problems through
proactive monitoring, trending and alerting on the performance of these critical business

4
processes. Addressing potential problems before the end user experiences any degradation in
application response time ensures a high level of service to the customer.

Service Level Tracking

Replaying critical business transactions effectively measures the response time from the
end user perspective. Using Quest’s transaction record and playback capabilities,
organizations can answer the critical question of “how long does it take an end user to
complete a business transaction?” This information is critical to any application
performance management solution because it not only alerts an organization of
application performance degradation, but also can be established as the cornerstone to
setting up application service level tracking. It is important to watch end user response
time, application performance, and availability service levels to ensure the application
is performing at its optimal level. Service level tracking also provides invaluable
information for reporting and trending, determining usage based patterns on workload
or time periods.

See expected response time versus what the actual end users are experiencing, along with collecting
valuable data related to the average response time, for trending.

Many solutions using active response time measurement provide measurement through
the use of prepackaged [from the software vendor] transactions to simulate end user
activity – often using simple business transactions such as logging into the application
and not simulating real world transactions. Quest allows organizations to record actual
business transactions from within the application and play them back to simulate the
activity of the end-user. Having the ability to record real business transactions that
closely simulate either typical and critical application activity provides an advantage
over using prepackaged transactions provided by other solutions. Once business
5
transactions are defined, they are set up for playback at defined intervals on a regular
basis. They then collect information for end user response time, allowing for
comparisons to the expected response time (service level). The old saying, “not all
users are created equal” illustrates the value of the ability to playback defined business
transaction. This capability provides a valuable application performance tool by
simulating transactions from geographic locations. Also, an end user from the corporate
location may experience a much better response time than remote users. Seeing
response time from several geographic locations allow organizations to troubleshoot
geographically when end-user response-time is degraded and pinpoint the performance
issue that services those users.

The ability to define steps in a business transaction is critical to end user response time.
Seeing exactly what part of the transaction is causing the end user experience to slow
down.

The most important goal of simulating business transactions is to ensure adequate end-
user response time. One benefit of this that is often overlooked is whether the
transaction provides a method for checking connectivity. With a traditional
management solution checking for faults of infrastructure, this does not typically ensure
that users can connect to the application. With the increase in complexity of today’s
applications, typically several components must be running in conjunction for the
application to actually be in an accessible state. Simply seeing that they are available
does not mean users can access the application. If an application goes into a hung or
confused state, an alert may never be sent to the organizations even though end users
may not be able to connect. By simulating a business transaction that logs on to the
application, the administrator will be notified immediately if a transaction is unable to
connect to the application.

6
Validating that end users are able to connect to the application.

Active response time measurement gives organizations a clear indication of application

performance and availability from the end user perspective.

Application Performance Management

The complete solution for Application Performance Management includes
encompassing not only end user response time but also detailed overall application
performance information. The ability to correlate the end user experience with the
application performance and the health of the underlying infrastructure gives the
business a complete perspective. This methodology provides the most complete way to
proactively detect, diagnose, and resolve application performance problems, allowing
organizations to go from end-user experience to application performance resolution in a
shorter period of time.

About the Author

Chris Spacko is a Product Marketing Manager for Monitoring at Quest Software, Inc. He has
over 10 years experience working with financial systems, ERP solutions and as a Software
Consultant. Chris has worked with Software Performance Solutions for over 5 years.

About Quest Software

Quest Software, Inc. (NASDAQ: QSFT) is a leading provider of application
management solutions. Quest provides customers with Application Confidencesm by
delivering reliable software products to develop, deploy, manage and maintain
enterprise applications without expensive downtime or business interruption. Targeting
7
high availability, monitoring, database management and Microsoft infrastructure
management, Quest products increase the performance and uptime of business-critical
applications and enable IT professionals to achieve more with fewer resources.
Headquartered in Irvine, Calif., Quest Software has offices around the globe and more
than 18,000 global customers, including 75% of the Fortune 500. For more information
on Quest Software, visit www.quest.com.

World Headquarters
8001 Irvine Center Drive
Irvine, CA 92618
www.quest.com
e-mail: info@quest.com
Inside U.S.: 1.800.306.9329
Outside U.S.: 1.949.754.8000
Please refer to our Web site for regional and international office information. For more
information on Quest Software’s Application Performance Management solutions or other
Quest Software solutions, visit www.quest.com.

Copyright © 2003 Quest Software, Inc. The information in this publication is furnished for information use only, does not
constitute a commitment from Quest Software Inc. of any features or functions discussed and is subject to change without notice.
Quest Software, Inc. assumes no responsibility or liability for any errors or inaccuracies that may appear in this publication.